EST 3590 is a prospective, randomized Phase III Eastern Cooperative Oncology Group trial in which patients with stage II or IIIa nonsmall cell lung cancer (NSCLC) undergoing a thoracotomy have a complete resection of their tumor and are then randomized to radiotherapy or chemotherapy plus radio-therapy. This high-priority intergroup study offers a unique opportunity to prospectively correlate the expression of biological or molecular markers, postulated to be relevant to human lung cancer, with patient stage and outcome in a defined group of patients, and to determine the relationship of these markers to each other within the same tumor sample. With the active and integral participation of thoracic surgeons from selected institutions on this trial, we expect to obtain tissue samples from a minimum of 150 patients over 3 years, providing a unique opportunity to prospectively determine the relevance of these molecular and cellular alterations to each other and to the outcome of patients with resectable NSCLC. The objectives of this proposal are to test the following hypotheses in primary, resected NSCLC tumors: (1) Activated K-ras and p53 mutations are negative prognostic factors in NSCLC and can be used in a prospective fashion to identify poor- prognosis patients with NSCLC. (2) Immunohistochemical staining can be used to screen for the mutant p53 oncoprotein. (3) Expression of blood-group antigen A is a favorable prognostic factor in NSCLC. (4) Markers of nuclear proliferation, tumor vascularity, and neuroendocrine differentiation, and the epidermal growth factor (EGF) receptor can also be correlated with tumor stage and outcome in NSCLC. (5) A multiple regression analysis of these factors may yield additional important prognostic information. The specific aims of this project are to: (1) establish a central repository at the University of Wisconsin Comprehensive Cancer Center; (2a) determine the incidence of K-ras mutations by differential hybridization using allele-specific oligonucleotide probes; (2b) identify p53 mutation by single-strand conformation polymorphism; putative p53 mutations will be characterized by nucleotide sequence analysis; (2c) determine the expression of the p53 oncoprotein by immunohistochemistry. (3a) assess group A blood antigen and EGF receptor levels; and (3b) evaluate neuroendocrine markers on keratin-negative, mucin-negative tumors by electron microscopy, and assess p105 nuclear antigen levels and Factor 8 levels by flow cytometric analysis; (4) correlate these results with histology, TNM stage, time to relapse, and survival, using multiple regression analysis.